Potential Selective Pressures by Parasitoids on a Plant-Herbivore Interaction Author(s): Arthur E. Weis and Warren G. Abrahamson Source: Ecology, Vol. 66, No. 4 (Aug., 1985), pp. 1261-1269 Published by: Ecological Society of America Stable URL: http://www.jstor.org/stable/1939179 Accessed: 30/01/2009 18:17 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=esa. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact
[email protected].
Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Ecology.
http://www.jstor.org
Ecology, 66(4), 1985, pp. 126 1-1269 t 1985 by the Ecological Society of America
POTENTIAL SELECTIVE PRESSURES BY PARASITOIDS ON A PLANT-HERBIVORE INTERACTION' ARTHUR E. WEIS2 AND WARREN G. ABRAHAMSON Department of Biology, Bucknell University, Lewisburg, Pennsylvania 17837 USA
Abstract. The potential role of the third trophic level in the evolution of plant-herbivore relationships was examined in the case of the goldenrod Solidago altissima, and the fly Eurosta solidaginis, which forms a round stem gall. Previous observation had shown that galls attacked by parasites are significantly smaller than those in which the gall maker survives. Two different mechanisms could cause such a pattern: parasite attack could occur before galls reach full size and the attacks could cause early cessation of growth, or attack could occur after galls reach their mature size, but with inherently small galls being more prone to attack. In the first instance, parasite attack would diminish the cost of the gall to the plant, and thus favor plant genotypes that facilitate parasites. In the second instance, parasites would exert selection pressure on the gall maker to induce larger galls. Monitoring of marked plants in the field, and field experiments in which parasites were excluded from gall-bearing plants except during controlled periods, showed that parasite attack does not stop gall growth. The parasitoid wasp Eurytoma gigantea is limited to attacking small galls because of the limited reach of its ovipositor. This created a selection intensity of 0.50, favoring Eurosta that induce larger galls. Evolutionary response to selection could be realized directly through change in the gall maker's stimulus ability, or less directly through phenological changes. Plant reactivity to the gall maker declined with plant age, so that late-starting galls were more vulnerable to parasite attack. Key words: coevolution; Eurosta; Eurytoma; gall makers; herbivore; natural selection; parasitism; Pennsylvania; phenology; Solidago; species interactions.
INTRODUCTION
Studies of plant-herbivore interactions have focused primarily on the direct effects that producer and consumer have upon the growth, survivorship, and reproduction of one another (for recent reviews see Thompson 1982, Crawley 1983, Futuyma 1983). However, direct effects may not be responsible for some of the adaptive traits involved in a species interaction. Species on higher trophic levels may exert selective pressures that cause evolution within an interaction between species on lower trophic levels (Bergman and Tingey 1980, Price et al. 1980). Influence by the third trophic level may be especially important to interactions between single plant species and monophagous, herbivorous insect species. The host plant is the habitat of the small, immobile, immature feeding stages of many insect species. Parasitoids and predators search for food within this "habitat," often using plant traits as cues (Vinson 1976). Herbivores may use structural features of the plant as refugia from attack, or plant morphology may simply impede search (Hulspas-Jordan and Van Lentren 1978, Price et al. 1980). Therefore, plant defenses against herbivory could include traits that facilitate attack by the herbivore's natural enemies. Conversely, herbivore defense against attack could include behavioral and physiological traits that take advantage of plant properties. The third trophic level may push I Manuscript received 25 April 1984; revised 24 September 1984; accepted 29 September 1984. 2 Present address: Department of Biological Sciences, Northern Illinois University, DeKalb, Illinois 60115 USA.
a plant-herbivore interaction in either direction; the direction taken will depend in large part on which of the interacting species is under stronger selection pressure. Here we present a study on the interaction between tall goldenrod, Solidago altissima (Compositae)3, and the gall-making fly Eurosta solidaginis (Diptera: Tephritidae). We examined the potential selective pressures exerted by the parasitic natural enemies of the gall maker to determine if they can be a significant force in the evolution of either the goldenrod's responsiveness to the gall maker or Eurosta's ability to stimulate gall growth in the plant. The experiments and field observations we report test the hypotheses that (1) parasite attack on the gall maker arrests gall development, thereby reducing the plant's cost of gall formation and (2) small galls, by virtue of their size, are more likely to be victims of successful parasite attack. Previous work on the goldenrod-gall fly interaction by Stinner and Abrahamson (1979) raised the possibility of a selective advantage to goldenrod individuals that facilitate parasitoid attack on the gall maker. These authors found that galls attacked by parasitoids were smaller and drew less energy away from normal plant growth processes. If parasitoids cause this reduced energy flow (by killing the 3This goldenrod is also known as Solidagocanadensisvar. scabra (Werner et al. 1980, Abrahamson et al. 1983). In previous papers (Hartnett and Abrahamson 1979, Stinner and Abrahamson 1979) we have referred to this species simply as
S. canadensis.
1262
ARTHUR E. WEIS AND WARREN G. ABRAHAMSON
gall maker) the plant should benefit, since galls lower the production of fruits and reduce vegetative propagation (Hartnett and Abrahamson 1979). Alternatively, galls that are inherently small may be more vulnerable to penetration by parasitoid ovipositors. This would result in a selective advantage to gall makers inducing large galls. In other systems variation in parasitoid attack frequency has been linked to variation in gall size and morphology (Askew 1961, Weis 1982, 1983); parasitoids may have exerted a major influence in the morphological evolution of cynipid wasp galls (Price 1980, Cornell 1983). The question of which, if either, selective regime can operate was engendered by another study, which showed that variation in gall size and growth rate is caused by genetic variance in both the goldenrod and the gall maker populations (A. E. Weis and W. G. Abrahamson, personal observation). Given heritable variation in gall traits, phenotypic evolution in either plant or insect is feasible if selective pressures are of sufficient magnitude.
Ecology, Vol. 66, No. 4
MATERIALS AND METHODS
This study was conducted in the summers of 1982 and 1983 at the Bucknell University Chillisquaque Natural Area, Montour County, Pennsylvania, USA. The host plant grew in two fields removed from agriculture within the past 10 yr. Phenology of the gall maker and its parasites
In 1983, daily emergence of gall maker adults was recorded from 1000 galls held in an outdoor screen cage. Median emergence date and first and third quartile dates, were determined from a plot of cumulative percent emergence vs. date. Emergence of the adult parasites was also monitored. The timing of gall appearance in 1983 was determined from a group of 719 shoots bearing oviposition scars. In late May, these were marked with numbered tags and even-numbered stems were examined for gall appearance at 3-4 d intervals. Median and quartile dates of appearance were calculated. A sample of 54 punctured buds was collected in the 1st wk of June in NATURAL HISTORY OF THE SYSTEM 1983 and all buds were dissected to determine the depth The larvae of Eurosta cause the growth of a spheroid- to which the egg had been inserted into the bud. Growth phenology of galls was determined from peshaped swelling on the stem of tall goldenrod. Although the related goldenrods S. canadensis (sensu strictu) and riodic destructive samples. In 1982, - 250 newly formed S. gigantea are infested by northern and western pop- galls were marked with numbered tags. At 3-7 d inulation of this insect, S. altissima is the sole host plant tervals, samples of 20 galls, chosen by a randomization in central Pennsylvania (Givens 1982). It is a perennial, computer program, were collected for measurement. clonal herb commonly found in old fields and along In 1983, the 360 odd-numbered marked stems were roadsides (Werner et al. 1980). sampled by the same procedure. We measured diamThe life history of the gall maker was detailed by eter at the equatorial plane. Galls were then cut in half Uhler (1951). The cycle begins in the spring when the at the equator, and the thickness of the gall wall was larva, which has overwintered in the previous season's measured at four points and averaged. In 1982, outer gall, pupates and then emerges to mate and oviposit. dimensions were measured with calipers and inner diIn mid-May the female injects an egg into the terminal mensions with an ocular micrometer. In 1983, a disbud through a tube-like ovipositor, leaving a scar. The secting microscope with a drawing tube attachment first-instar larva tunnels through the bud and into the was used to superimpose the image of the gall on an stem just below the apical meristem. Newly formed APPLE II graphics tablet; the image was traced with galls become apparent by mid-June and reach their a magnetic pen, and a computer program calculated mature diameter within 3 wk. The larvae feed on the the dimensions from the tracing. inner tissues of the gall until October, when they enter Phenology of enemy attack was determined by a diapause. serial exclusion/exposure experiment. In each of the The primary natural enemies of Eurosta during its two years - 700 galls were marked with numbered tags larval growth stage are the parasitoid wasps Eurytoma as soon as they became visible. Fifty galls (60 in 1983) obtusiventris and E. gigantea (Hymenoptera: Eurytom- were left unmanipulated as controls. The remaining idae) and the inquiline beetle Mordellistena unicolor galls were covered with parasitoid exclusion sleeves of (Coleoptera: Mordellidae) (Uhler 1951, 1961, Cane and nylon organdy (5 x 12 cm) secured with wire ties and Kurczewski 1976, Stinner and Abrahamson 1979, paper clips. Immediately, a randomly chosen subsamAbrahamson et al. 1983). E. obtusiventris is an internal ple of 50-60 galls was uncovered for 1 wk, during which parasitoid that delays development until September. time parasites were free to attack. At week's end galls By contrast, E. gigantea lays its egg within the gall's were re-bagged and the next subsample was exposed. central cavity; the larva feeds externally on the gall In each year, one additional group remained bagged maker larva, then consumes gall tissue (Uhler 1951). for the entire season to test sleeve efficacy. At the seaThe beetle lays its egg just below the gall's epidermis; son's end the galls were dissected and their contents the larva then tunnels through the gall, eventually end- identified. Comparison of mean diameters of the exing in the central chamber, where it consumes the gall posure groups to an unbagged control group showed maker. that bagging had no effect on gall size.
August 1985
PARASITES AND PLANT-HERBIVORE EVOLUTION
1263
us to suspect the diminished growth of late-starting galls to be due in part to clonal (genotypic) differences in plant reactivity rather than to plant age per se. AlES Yternatively, "clusters" of galls could be the progeny of gall individual female flies, with each cluster inheriting genes for fast or slow gall formation. These possibilities were investigated in a greenhouse experiment in which plant genotype and age were crossed factors, and fly genotype MU was randomized. Plants were grown in 15-cm pots from 2-g rhizome cuttings taken from four S. altissima clones growing near Lewisburg, Pennsylvania. Three groups EG were started at 12-d intervals starting on 1 March 1983. 9~Eight weeks after the first group was started, all groups 3 1 27 18 2 19 5 13 were exposed to Eurosta oviposition by placing a plas----MayJune July---tic sleeve over the plant and introducing a randomly 1983 DATE OF EMERGENCE selected, mated female. All 108 shoots were punctured and 45 formed galls. Gall diameter was measured at FIG. 1. Dates of emergenceof Eurosta and its parasites and of the appearanceof new galls.Overwinteredinsectswere four dates. Data were analyzed in a factorial, repeatedreared from their galls in an outdoor emergencecage; gall measure analysis of variance after logarithmic transformation was monitored on marked plants. Vertical lines formation to equalize variances. indicatemedians,horizontallines show ranges,and wide horizontal bars connect the first and third quartiles.Abbreviations (sample sizes in parentheses):ES = Eurostasolidaginis 30 (92 86, 94 YY, 195 galls); EO = Eurytoma obtusiventris (14
1982
individuals);MU = Mordellistenaunicolor(66 individuals); EG = Eurytoma gigantea (17 d6; 31 YY).
The exclusion/exposure experiment in 1983 was also used to determine if parasitoid attack stops gall growth. Gall diameters were measured at the beginning of each exposure interval, then again at the time of dissection. Mean postexposure growth increments of E. giganteaattacked and unattacked galls were compared in each exposure group. Plant age at gall appearance: effect on growth and survivorship We tested whether plant age at the time of gall appearance has an effect on gall size by following the growth of age cohorts in the field. In 1982, 150 punctured stems were marked in late May; 359 stems (the even-numbered plants described above) were used in 1983. All stems were censused at 2-4 d intervals throughout June, then at 4-7 d intervals into August. Cohorts consisted of the set of galls first discovered on a census date. Diameters were measured on all galls on every census date. Final cohort mean diameters were compared with one-way analysis of variance and the Student-Newman-Keuls test for separation of means. In October of each year, the galls were dissected to determine if the incidence of parasite attack varied with the date of gall appearance. In order to best meet the requirement of minimum expected values for a chisquare test, cohorts were pooled into early and late groups. Small sample sizes dictated the use of Fisher's Exact Test in some instances. In both years, it seemed many of the galls on any one plant clone appeared on the same date. This led
20
*-S
diameter
0--
wall
A---V&
0~
larva
10
Z
00
_
.
/4F
a
Lj] 3 0-
1983
E
0 20-
10
00
-A-A---
10
20
JUNE
301
10
JULY
20
10
311
AUG
FIG. 2. Growth curves for 0 gall diameter, 0 gall wall thickness,and A Eurosta larval length. On each date 15-25 galls werecollected,dissected,and measured.Points indicate means, and vertical lines indicate ? 1 SD. Curveswere fitted by eye.
ARTHUR E. WEIS AND WARREN G. ABRAHAMSON
1264 10- E.obtusiventris
O.
Ecology, Vol. 66, No. 4
RESULTS
.0
Phenology of insect emergence and gall appearance
